Horsepower Thoughts

I’m in the midst of reading (a thus-far very good) review of the beta release of PHP5.

Hmm…a review of a preview….

One thing caught my attention: PHP5 comes with a built-in database, a lite (less-filling, tastes great) SQL database aptly named SQLite. While the author really didn’t come up with any good reason for PHP including this as part of the build, he did say it was not meant as a replacement for mySQL or Postgres.

OK.

But then he goes on and says it might be worth using for some tasks, such as handling sessions or what have you.

OK again, even though I don’t get running two databases against one app.

But his reasoning for this is that SQLite is fast, about twice as fast as mySQL (no slacker in the speed department).

OK – what’s going on here?

I thought we were pretty much past the point of the limitations of software speed because the hardware has gotten so damn fast.

While the Amazons and Googles of the world will still be buying faster and multi-processored (is that a word?) boxes, the rest of us really don’t have a need for speed. Either on the sites we develop or our home machines.

Sure, faster is always better, but it has gotten to the point were there are diminishing returns.

The leap from a 486 to a Pentium was huge, and the Pentium Pro (essentially the basis for P2s, without the MMX or whatever) was a big jump in stability and robustness. (Hell, my main Linux box here is running on an approximately 6-year old Pentium Pro 200Mhz machine. It was my Windoze box for three years.)

But how much of a jump – really – is there from P3 to P4? Sure, the numbers are going up, as are the buses (way underappreciated by buyers, who look at Mhz only), but how much faster – especially for the average user – is a 3.0Ghz box than a 2.0Ghz box?

Sure, it’s faster, but noticably?

We are really past that. I’m working on a 1Ghz P3 right now (which was a screamer after my 200Mhz Pentium Pro), and it’s still a fast box.

Sure, it would be nice if Photoshop started up faster, but that’s more of an OS issue (I have the same program on a Win98 box, half as fast, one-quarter the RAM and it’s faster there).

Right now, I’m running the following stuff – as services, not talking about programs I’m using (IE, command shell…):

  • MS SQL Server 7
  • mySQL
  • Netscrape FastTrack server (three instances of it)
  • IIS (v5)
  • Cold Fusion Server (v5, with the Graphing server, as well)
  • SMTP

That’s just the stuff I remember. But that’s stuff that runs in the background all the time, not even counting the three versions of IE, telnet, FTP, Outlook, Netscrape and other crap that I have open. And not to mention Norton (which is scanning as I type; then you can notice a little hang in programs opening, because Norton is close to pegging the CPU).

And nothing really hangs, all is pretty snappy.

And this is – for this day and age – a pretty low-end machine, and I run it hard.

So what’s this “oh goodie, it’s faster than mySQL” stuff? So a 100 row select on three joins will take 1/2 of a millisecond instead of a full millisecond? Oh goodie..

To be fair, people using PHP are running sites, and that is really where you want to shave the milliseconds, so there is some method to this madness (of a way of thinking…).

But I thought we were getting beyond this need for speed with quad-processor, gigs of memory boxes. Again, faster is always better, but – again – is the faster really noticable?

And – in this case (SQLite) – is the speed improvement enough do decide to split your app so parts (say, sessions) run against one database, other parts against another DB?

Ack! I say.

Proud to be an American?

I try to keep this blog on technical or whimsical issues, but sometimes….

OK, the Americans have just killed two of Saddam Hussein’s sons – Odai and Qusai.

And this is cause for much celebration in the US of A.

Why?

Because we are, essentially, assassins? (United States of Assassins?)

We invaded them.

They did not provoke us. They were belligerent, to be sure, and probably the country (Iraq) is better off without them, but…come on.

Again: We invaded them. Over the objections of almost all the rest of the world. On – at best – not the best of evidence of evil.

We put a price on their heads, their faces on “most wanted” playing cards. What are we, the Sopranos?

And yeah yeah yeah…things will be better without Saddam and sons, but I guess that’s the old “the ends justify the means” speech.

And which “end”? – tomorrow, a year from now, five years, 100 years?

Because things are not better for most average citizens there right now – war damage, looting, lack of power and water due to war damage.

And the desire to run their own country, which the US does not seem to want to do, at least right away (right or wrong). So the “end” of “now” is not a justification.

Give me a ring when it is.

And yes, I know John Ashcroft is listening…..

RSS Fears Eased

There are about a million comments I want to make about Dave Winer’s decision to relinquish ownership of RSS 2.0 to a non-profit group, but let this suffice:

This is a very good thing, and a very good gesture by Dave, who – to be honest – has a lot invested in RSS, has a lot to lose by releasing it, and a lot to gain by releasing it.

History will tell, but this is a good move for damn near everyone.

I started tuning out the RSS flak about two months ago; the rhetoric … well, it was rhetoric.

Don’t need that.

Need specs. Standards. Blah blah blah…

This will help.

Like/dislike/indifferent to Dave, this is a generous and most excellent move.

Thanks, Dave.

What Broadband Means

I was just reading an article (posted a couple days ago) on news.com about how the Bells are hindering US broadband expansion.

Read the article yourself for the opinion – basically, handing the keys to control the broadband doors back to the incumbent Bells will slow broadband use in the US. OK, I agree with the author in many ways, but the article set me off tangentially.

(Don’t you just love it when you’re tangential?)

I was thinking about how I’ve seen, oh, five billion or so ads [print and Web]/TV commercials and so on touting broadband and what it means.

And in most cases, what was really touted as the big deal was the “rich media” experience.

Flash tools/widgets/games.

Streaming news/entertainment clips.

Net conferencing.

Stuff like that – that, before broadband, you could only do on your computer, because even the slowest computer had a much faster CPU/bus combo than 2400-5600 baud. You could only squeeze so much out of that copper before DSL came along.

OK. With broadband (DSL, cable, Dish, I don’t care….) you can do a lot of the stuff as advertised via the Net.

But that misses the whole point.

The outstanding things about broadband are the following two points (yes, only two):

  • Always on. With a mature OS (be it a UNIX variant or Win2000; let’s not quibble), you can leave the computer on for extended (week plus) periods and always have the Internet an application launch away (browser, e-mail, whatever). THIS is the broadband killer app. Example: I can’t tell you how many times I’ve been watching a movie/TV with someone and gotten into a discussion like “Wasn’t [he/she] in?…” or whatever. So I hit Google or nbc.com or whatever. I get the answers; advertisers get the impression. I would never boot up just for this. NEVER.
  • Fast for ordinary tasks. While a distant second in importance, this is a broadband feature that is way undersold. Yes, Yahoo! and Google are optimized to be fast on even the slowest connections. But what if you want to shop at baitAndTackle.com? Think they’ve optimized the site – with the 50K images and embedded WAV files – for narrowband? Nah… The ability to do the normal without a wait is HUGE.

Ever wonder why AOL is losing money and customers? MSN is fighting to get customers?

Well, there are about a million reasons, but one of the things that always amuses (i.e. laugh or you’ll cry) me is how much money is tossed at building these so-called “Broadband Portals.”

Who gives a rat’s ass?

While many people – the timid or newbies – will behave like the Intenet begins and ends with the portal and its proprietary offerings, most have either outgrown this “BBS-like” system or at least get the concept that there is something outside the AOL garden.

Google doesn’t advertise. Everyone uses it; it makes money.

News.com and Slashdot have huge followings. They are not portals.

Ebay and Amazon are totally doable on narrowband (but, obviously to me better on broadband. Just for the speed in and out). But on broadband, I can skate through the sites more quickly. Yes, I may miss some things I may have seen had I had to wait for a page to roll in, but I think – overall – I see more when the pages snap in at near T-1 speeds.

Broadband is good for business: Advertisers and sellers.

Buyers, as well.

Broadband is all about removing the mystique of the Internet (or, at least, the Web), of making it possible to quickly navigate (hence, find more, stumble upon more), of allowing people to move outside the walled gardens of AOL/MSN etc.

This is a good thing. It’s the first step in making the Internet transparent to most users.

Think about it – think about your mom/grandmother – booting up (“booting up?”) a computer and listening to the modem trying to connect with it’s screeches and whines. And frequent and unexpected drops.

The techie gets it; we are part of a very small minority overall.

Isn’t it better to have a computer like a radio or TV (better, because you can put it in sleep mode) that is always just a click away from the Internet. That collects e-mail while you sleep?

Why is that bad?

Why isn’t that advertised more heavily?

Archive is Back

And for some inexplicable reason, Blogger has decided to publish – and publish correctly – my archive page.

Go figure.

I think I’ll keep mine instead, just for pride…I hacked it together in a short time because I needed it.

Archive Fix

Well, my notes to Blogger Control didn’t help, and the problem with my archives just kept getting worse – first, the archive index was not properly parsed, then (today) it just wouldn’t write at all – so I came up with my own solution.

Thanks to the magic of Perl, I just built a dynamic page that scans the directory and writes the archive links. (Archive Index)

Hey, the damn thing even works…

I understand that Bloggle is growing/going through the transition phases of its integration (Blogger => Google), and they have done some good work. The actual browser-based tool that is used it a vast improvement over the old tool (perfect? no. but what fun would that be?), and the publishing seems way faster – probably due to moving to Google’s more stable and scalable platforms.

That doesn’t mean that I liked having to craft my own solution, but I understand the transition issues and, hell, this is currently a free service.

Fair is fair.

And it was fun to quickly whack together the Perl solution.

As I’ve mentioned before, I seem to be turning to Perl more and more for quick solutions, simply because it is so damn flexible.

Bets on when I get my Blogger archive back??

Still No Archives

Still no archive index.

The actual archive pages are generating OK, but the index is not.

Blogger has a note about this in support, but the fix outlined did not work for me.

Oh well, at least the entries do exist; and it’s really no matter – no one reads this blog, anyway.

Worst case scenario, I’ll write a Perl script to create the index on my server if this doesn’t get fixed quickly.

Bad Woojoo

OK, this is not good.

With this “new” Blogger (very nicely done, for the most part), my site is now hosed.

I republish the entire site and the archives don’t come back.

Ouch.

Hmm…I don’t see anything in the status.blogger.com site…

The archive pages are actually there, but the index page is hosed – it does not do the list; it looks like the page was not parsed before publishing. It comes across with the blogger tags – and only one line (therefore, not looping as necessary).